基于传声器阵列的旋转声源识别方法研究

发布时间：2018-05-24 05:14

本文选题：旋转声源识别 + 多普勒效应　；参考：《合肥工业大学》2017年硕士论文

【摘要】：以叶轮机械为代表的旋转机械在给工业生产提供动力的同时,也产生了严重的噪声污染问题。降低这些旋转机械的噪声迫在眉睫,而控制声源是降低噪声的最根本措施。基于传声器阵列识别声源可为声源的主被动控制提供重要的指导,然而目前的声源识别方法主要是针对静止声源。由于旋转声源所产生的声信号中包含明显的多普勒效应,因此现有方法在识别旋转声源时会失效。围绕该问题,本论文将研究基于传声器阵列的旋转声源识别方法,为旋转声源识别提供理论依据和指导方法。具体研究内容如下:第一章首先阐述了旋转声源识别的背景和意义,然后通过论述旋转声源识别技术的研究现状,并分析其中存在的问题,提出本论文的研究内容。第二章针对现有旋转单极子声源识别方法计算效率低的问题,提出了基于DAMAS2的自由场旋转单极子声源识别方法。首先分析了自由场中旋转单极子产生的多普勒效应,以及其对声源识别结果的影响;然后采用旋转框架技术消除多普勒效应的影响;继而采用DAMAS2解卷算法识别旋转声源,提高声源识别分辨率和计算效率;最后,通过数值仿真和实验分析验证了所提基于DAMAS2的自由场旋转单极子声源识别方法的有效性。第三章针对现有旋转声源识别方法无法准确识别实际工程中常遇到的旋转偶极子声源问题,提出了自由场中旋转偶极子声源识别方法。考虑到实际旋转机械产生的声源中,周向偶极子和轴向偶极子占主要部分,因此基于球谐波展开的方法推导出这两种偶极子Green函数,并分析这两种偶极子Green函数在实际应用过程中存在的模态截断数的选取问题。通过引入旋转框架技术消除多普勒效应后,在高频采用波束形成技术来识别声源,在中低频采用反技术来识别声源。最后通过数值仿真和无人机旋转桨叶的实验验证了所提方法的有效性。第四章针对现有管道内旋转声源识别方法测量成本高的问题,提出了基于正交匹配追踪算法的管道内旋转声源识别方法。首先分析了管道内旋转声源的辐射声场特性;然后采用旋转框架技术消除了管道内旋转声源产生的多普勒效应;之后采用基于正交匹配追踪算法的方法来实现管道内旋转声源识别,以达到利用较少传声器测量来获得高分辨率声源识别的目的,进而减少测量成本、提高测量效率;最后,通过数值仿真验证了所提方法的有效性和优越性。第五章总结了本文的主要研究成果,提出了需要进一步研究和解决的问题。
[Abstract]:The rotating machinery, represented by impeller machinery, not only provides power for industrial production, but also causes serious noise pollution. It is urgent to reduce the noise of these rotating machinery, and controlling the sound source is the most fundamental measure to reduce the noise. The recognition of sound sources based on microphone arrays can provide important guidance for active and passive control of sound sources. However, the current methods of acoustic source identification are mainly aimed at static sound sources. Due to the obvious Doppler effect in the sound signal generated by the rotating sound source, the existing methods will fail in identifying the rotating sound source. Around this problem, this paper will study the method of rotating sound source recognition based on microphone array, which provides theoretical basis and guidance method for rotary sound source recognition. The specific research contents are as follows: in the first chapter, the background and significance of rotary sound source recognition are introduced, and then the research status of rotary sound source recognition technology is discussed, and the existing problems are analyzed, and the research content of this paper is put forward. In the second chapter, aiming at the problem of low computational efficiency of the existing methods for the recognition of rotating monopole sound sources, a free-field rotating monopole source recognition method based on DAMAS2 is proposed. Firstly, the Doppler effect of rotating monopole in free field and its influence on the result of sound source identification are analyzed. Then, the influence of Doppler effect is eliminated by rotating frame technique, and then the rotating sound source is identified by DAMAS2 unwinding algorithm. Finally, the effectiveness of the proposed method based on DAMAS2 is verified by numerical simulation and experimental analysis. In the third chapter, aiming at the problem that the existing methods of rotary sound source identification can not accurately identify the rotating dipole sound source which is often encountered in practical engineering, a method for the identification of rotating dipole sound source in free field is proposed. Considering that the circumferential dipole and the axial dipole are the main parts of the sound source produced by the actual rotating machinery, the Green functions of the two dipoles are derived based on the spherical harmonic expansion method. The selection of modal truncation in the practical application of the two dipole Green functions is analyzed. After the Doppler effect is eliminated by introducing the rotating frame technique, the beamforming technique is used to identify the sound source at high frequency and the inverse technique is used to identify the sound source in the middle and low frequency. Finally, the effectiveness of the proposed method is verified by numerical simulation and experiments on rotating blades of UAV. In chapter 4, aiming at the problem of high measurement cost of the existing methods for identifying rotating sound sources in pipelines, a method based on orthogonal matching tracking algorithm is proposed for identification of rotating sound sources in pipelines. Firstly, the characteristics of the radiated sound field of the rotating sound source in the pipeline are analyzed, and then the Doppler effect caused by the rotating sound source in the pipeline is eliminated by using the rotating frame technique. Then the method based on orthogonal matching tracking algorithm is used to realize the identification of rotating sound source in pipeline, so as to achieve the purpose of obtaining high resolution sound source identification by using fewer microphone measurements, thus reducing the cost of measurement and improving the measurement efficiency. Finally, The effectiveness and superiority of the proposed method are verified by numerical simulation. The fifth chapter summarizes the main research results of this paper, and points out the problems that need to be further studied and solved.
【学位授予单位】：合肥工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TB53

【相似文献】